Durable client-side caching for distributed storage

Patent metadata

Field	Value
Publication number	US-10895997-B2
Application number	US-201916281044-A
Country	US
Kind code	B2
Filing date	Feb 20, 2019
Priority date	Feb 20, 2019
Publication date	Jan 19, 2021
Grant date	Jan 19, 2021

Abstract

Systems and methods are disclosed herein for writing data, including storing, by a storage client, a cache copy of the data in a memory of the storage client and receiving, by the storage client, acknowledgement indicating that a single replica of the data is made in a storage node remote to the storage client.

First claim

What is claimed is:

1. A non-transitory computer readable media includes computer-executable instructions embodied thereon that, when executed by a processor, cause the processor to:
store a cache copy of a data in a memory of a storage client;
receive write acknowledgement indicating that a single replica of the data is made in a storage node remote to the storage client;
determine that the memory is to be freed responsive to determining a workload on the storage client associated with the cache copy is being moved to another node;
send the data to another storage node for making an additional replica of the data in response to determining that the memory is to be freed.

2. The non-transitory computer readable media of claim 1 , wherein the cache copy is a durable replica of the data.

3. The non-transitory computer readable media of claim 1 , wherein the cache copy is made at a cache layer of the memory of the storage client.

4. The non-transitory computer readable media of claim 1 , wherein the processor is further caused to free a portion of the memory storing the cache copy.

5. The non-transitory computer readable media of claim 1 , wherein the storage client is a compute node.

6. The non-transitory computer readable media of claim 1 , wherein the processor is further caused to send the data to the storage node for making the single replica of the data.

7. The non-transitory computer readable media of claim 1 , wherein the processor is further caused to
receive another acknowledgement indicating that the additional replica of the data is made in the another storage node.

8. The non-transitory computer readable media of claim 7 , wherein the memory is freed responsive to receiving the another acknowledgement.

9. The non-transitory computer readable media of claim 1 , wherein the processor determines that the memory is to be freed responsive to further determining additional space is needed in the memory.

10. The non-transitory computer readable media of claim 1 , wherein
the acknowledgement is configured to notify the storage client that a replica of the data has been made in a storage backend of the storage client;
the storage backend is connected to the storage client via a network; and
the storage backend comprises the storage node.

11. A storage client comprising:
a memory; and
a processor with programmed instructions to:
store a cache copy of data in the memory;
receive an acknowledgement indicating a single replica of the data is made in a storage node remote to the storage client;
determine that the memory is to be freed responsive to determining a workload on the storage client associated with the cache copy is being moved to another node;
send the data to another storage node for making an additional replica of the data in response to determining that the memory is to be freed.

12. The storage client of claim 11 , wherein the cache copy is a durable replica of the data.

13. The storage client of claim 11 , wherein the cache copy is made at a cache layer of the memory.

14. The storage client of claim 11 , wherein the processor has further programmed instructions to free a portion of the memory storing the cache copy.

15. The storage client of claim 11 , wherein the processor has further programmed instructions to send the data to the storage node for making the single replica of the data.

16. The storage client of claim 11 , wherein the processor has further programmed instructions to
receive another acknowledgement indicating that the additional replica of the data is made in the another storage node.

17. The storage client of claim 16 , wherein the memory is freed responsive to receiving the another acknowledgement.

18. The storage client of claim 11 , wherein the memory is a portion of memory in a flash memory.

19. A method comprising:
storing, by a storage client, a cache copy of a data in a memory of the storage client;
receiving, by the storage client, acknowledgement indicating that a single replica of the data is made in a storage node remote to the storage client;
determining, by the storage client, that the memory is to be freed responsive to determining a workload on the storage client associated with the cache copy is being moved to another node; and
sending, by the storage client, the data to another storage node for making an additional replica of the data in response to determining that the memory is to be freed.

20. The method of claim 19 , further comprising sending the data to the storage node for making the single replica of the data.

21. The method of claim 19 , further comprising receiving another acknowledgement indicating that the additional replica of the data is made in the another storage node.

22. The method of claim 21 , wherein the memory is freed responsive to receiving the another acknowledgement.

23. The method of claim 19 , wherein the memory to be freed is a portion of a non-volatile memory local to the storage client.

24. The method of claim 19 , wherein the cache copy is a durable replica of the data.

25. The method of claim 19 , wherein the cache copy is made at a cache layer of the memory.

26. The method of claim 19 , wherein the storage client is a compute node.

Assignees

• Nutanix Inc

Inventors

• Franciosi Felipe
• Cui Miao

Classifications

• G06F2212/151

  Emulated environment, e.g. virtual machine · CPC title

• G06F12/0866

  for peripheral storage systems, e.g. disk cache · CPC title

• H04L67/1095

  Replication or mirroring of data, e.g. scheduling or transport for data synchronisation between network nodes · CPC title

• H04L67/1097 Primary

  for distributed storage of data in networks, e.g. transport arrangements for network file system [NFS], storage area networks [SAN] or network attached storage [NAS] · CPC title

• G06F3/0647

  Migration mechanisms · CPC title